
Eating disorders have traditionally been grouped according to what patients do. A person restricts food. Another binges. Someone purges. Still others exercise compulsively. A patient becomes dangerously underweight. Another lives in a larger body while experiencing episodes of uncontrollable eating. Clinicians understandably focus on these behaviors because they are the immediate manifestations of the illness and, in some cases, can become life-threatening. Now, genetics is beginning to help us understand some of the biology that may help explain why those patterns develop.
A major international genetic study published on August 19, 2026, in Nature Mental Health analyzed genetic data involving more than 24,000 people with anorexia nervosa and more than 39,000 people identified as having binge-eating behavior across several datasets, together with more than one million comparison participants in each analysis. Researchers from around the world participated, including investigators from the University of Otago in Christchurch, New Zealand.
That study can be found here:
https://www.nature.com/articles/s44220-026-00698-2.pdf?utm_source=chatgpt.com
The findings do not give us a genetic test for an eating disorder, identify an “anorexia gene” or “binge-eating gene,” or establish that environment, trauma, dieting, culture or psychological experiences are unimportant.
What the findings do reveal may be more consequential. The study strengthens the evidence that anorexia and binge-eating behavior involve numerous genetic influences and that these seemingly different presentations can share important underlying vulnerabilities while diverging substantially in other areas of biology. Perhaps most importantly, the findings provide additional evidence that eating disorder biology cannot simply be reduced to the biology of being thin, being heavier or having a particular body mass index. That distinction could eventually change how eating disorders are understood, classified and treated.
What Does a Genetic Study Like This Actually Tell Us?
The study is what scientists call a genome-wide association study, or GWAS. The concept is easier to understand than the terminology. Human beings share almost all their DNA, but millions of small genetic differences exist among us. Researchers can compare very large groups of people with and without a particular illness or behavior and ask whether certain genetic variations appear more frequently in one group than another. With sufficiently large numbers, patterns begin to emerge.
Researchers generally are not finding one gene that causes an eating disorder. Eating disorders appear to be polygenic, meaning many genetic variations, each usually contributing only a very small amount of risk, interact with one another and with a person’s biology, development and environment. A useful way to think about genetic vulnerability is as hundreds or thousands of tiny weights placed on a scale. No single weight determines the outcome, but together they can influence susceptibility. That is why someone can carry substantial genetic risk and never develop an eating disorder, while another person with genetic vulnerability may develop one when that vulnerability interacts with other biological, psychological or environmental factors. Genes are not destiny, but neither are they irrelevant background information.
Anorexia and Binge Eating Share Something Important
One of the study’s most interesting findings is that anorexia nervosa and binge-eating behavior share a meaningful amount of genetic vulnerability. At first, that may seem surprising. Anorexia and binge eating can appear almost opposite. One may involve severe restriction, while the other involves episodes of eating accompanied by loss of control. Anorexia is often associated with very low weight, while binge eating occurs across the weight spectrum and is frequently associated with higher weight. Yet their genetics suggest that they are not completely unrelated.
The researchers found a moderate genetic relationship between anorexia and binge eating, while additional modeling suggested substantial overlap among genetic variants contributing to susceptibility to the two phenotypes. At the same time, important differences emerged. Binge eating showed stronger genetic relationships with ADHD, smoking, problematic alcohol use and general risk tolerance. Anorexia showed a different pattern, including a stronger genetic relationship with obsessive-compulsive disorder and the unusual combination seen in previous research of psychiatric, metabolic and body-composition associations.
These findings must be interpreted carefully. They do not mean that binge eating is simply a problem of impulse control or that everyone who binges has ADHD, abuses alcohol or engages in risky behavior. Nor do they mean that everyone with anorexia has OCD. These are population-level genetic relationships suggesting that some biological systems involved in these other traits may overlap with systems affecting susceptibility to eating disorder behaviors.
The findings may also help explain something clinicians routinely observe. Eating disorders do not always remain neatly within diagnostic boundaries. A person may severely restrict at one stage of illness and later develop binge or purge symptoms. Another may move between diagnoses over time. Binge eating itself can occur in binge-eating disorder, bulimia nervosa and some presentations of anorexia nervosa. At least some of that clinical overlap may eventually prove to reflect shared biology rather than simply changing behavior.
Eating Disorders Are Not Simply Disorders of Body Weight
One of the study’s most important analyses involved BMI. Because anorexia is frequently associated with very low body weight and binge eating is often associated with higher weight, the researchers asked whether their genetic findings were essentially detecting genes related to body size rather than genes related to the eating-disorder phenotypes themselves. The answer was largely no.
For binge-eating behavior, the genetic component shared with BMI accounted for about 12 percent of the genetic variation being modeled, while approximately 88 percent remained in a component not shared with BMI. For anorexia nervosa, the BMI-shared component accounted for approximately 10 percent, while about 90 percent remained in the non-BMI component.
Those numbers require careful explanation. They do not mean that 90 percent of anorexia is genetic, that weight is unimportant in anorexia, or that 90 percent of the causes of anorexia have nothing to do with body weight. They mean something much more specific: when researchers examined the genetic signals associated with anorexia, most could not simply be explained by genetics associated with having a lower BMI. The same principle applied to binge eating and higher BMI.
The researchers concluded that their findings were not simply “BMI studies by proxy.” That is a significant distinction because weight can be a critically important manifestation of an eating disorder without being an adequate explanation for the disease itself.
Being Naturally Thin Is Not the Same Thing as Having Anorexia
Another finding reinforces that point. Researchers compared anorexia with persistent thinness, meaning people who remain naturally very thin but do not have anorexia nervosa, and found no significant genetic overlap between persistent thinness and anorexia.
Clinically, this may seem obvious, but biologically it matters. Imagine two people with similarly low body weights. One eats normally, does not fear weight gain, does not compulsively restrict and has always been naturally thin. The other has anorexia nervosa and experiences restriction, fear, rigidity, compulsive behaviors or other features of the illness. Their bodies may look similar, but that does not mean the underlying biology is the same.
The finding also reinforces why clinicians should be cautious about defining restrictive eating disorders by weight alone. A person with atypical anorexia nervosa can experience severe restriction, psychological symptoms and significant medical compromise without ever reaching a conventionally underweight BMI. The disease is not simply the number on the scale.
Anorexia Is Looking Increasingly Like an Illness of Both the Brain and the Body
This study builds upon an important finding from major anorexia genetics research published in 2019. That earlier work found genetic relationships not only between anorexia and psychiatric conditions, but also with metabolic characteristics, physical activity and body composition. The researchers proposed reconsidering anorexia as a “metabo-psychiatric” disorder, meaning an illness involving both psychiatric and metabolic biology.
The new study strengthens that broader concept. Our traditional understanding of anorexia can become too linear: a person develops psychological problems, those problems cause restriction, restriction causes starvation and starvation then causes metabolic abnormalities. Some of that unquestionably occurs because starvation profoundly affects virtually every organ system, including the brain.
However, genetics raises another possibility. Some metabolic characteristics may contribute to a person’s vulnerability to anorexia rather than existing only as consequences of starvation. That does not mean metabolism causes anorexia. It means that brain function, behavior, appetite, energy regulation, activity and metabolism may be considerably more intertwined than our traditional categories have suggested. If that proves correct, the implications for treatment could eventually be substantial.
Why Can Recovery Be So Difficult?
Families often encounter one of the most painful mysteries of eating disorders. A person may understand intellectually that the illness is dangerous, know that they need nutrition and desperately want their life back, yet remain unable to eat adequately, tolerate weight restoration, stop exercising or resist the behaviors of the illness. From outside the disease, this can be extraordinarily difficult to understand.
Genetic research does not yet tell us exactly why this happens, but it gives us another reason to reject simplistic explanations based on stubbornness, vanity, defiance or lack of motivation. If eating disorder vulnerability involves biological systems affecting anxiety, compulsivity, reward, appetite, activity, metabolism or behavioral control, then a patient may be experiencing pressures that neither the patient nor the clinician can currently measure. A person can rationally understand something and still have powerful biological systems pushing in another direction.
The emerging genetics do not yet explain treatment resistance or recovery at the individual level. What they provide is a biological framework within which researchers may eventually understand these clinical experiences much better.
The Future May Be Less About “Which Eating Disorder?” and More About “Which Version?”
This may be the most important long-term implication of the research. Today, diagnosis is based primarily on symptoms. Two patients who meet diagnostic criteria for anorexia nervosa receive the same diagnosis, yet those two people may be remarkably different. One may have pronounced anxiety and obsessive-compulsive characteristics. Another may have an unusually strong drive for physical activity. Another may have prominent metabolic features, while another repeatedly moves between restriction and binge or purge behaviors.
What if some of those differences reflect more than personality or presentation? What if they represent different biological pathways leading to what we currently call the same disorder? This study does not identify such biological subtypes today, but it points toward a research direction in which they may eventually become identifiable. The researchers themselves raise the possibility that genetics could someday contribute to a more biologically informed classification of eating disorders.
Medicine has undergone similar transformations elsewhere. Cancer provides an imperfect but useful comparison. Modern oncology increasingly examines molecular differences among tumors because two people with cancer in the same organ may have diseases that respond very differently to treatment. Eating disorders are nowhere near that level of precision, but the same conceptual possibility is important.
Imagine a future in which two patients with anorexia are no longer automatically assumed to have biologically identical illnesses simply because they satisfy the same diagnostic checklist. Treatment might eventually consider which biological systems appear most involved in each person’s disease. That is the promise of precision medicine.
Such a change could affect psychotherapy, medication development and clinical trials. A patient whose illness is strongly associated with rigidity and compulsivity may eventually benefit from a different therapeutic emphasis than someone whose dominant vulnerability involves reward processing or emotional regulation. Researchers testing medications may also become better able to identify subgroups who respond to treatments that appear ineffective when everyone with the same diagnosis is analyzed together. These are possibilities, not findings established by this study, but they are reasonable directions suggested by a growing body of biological research.
Could Genetics Lead to New Medications?
Potentially, although not immediately. The study identified several regions of DNA associated with anorexia and binge-eating behavior. For binge eating, one of the strongest associations involved the region around FTO, a gene long connected with body weight. The researchers raised the possibility that part of FTO’s relationship with higher BMI could involve eating behavior itself.
For anorexia, researchers confirmed several previously identified regions and identified additional ones. They also found evidence involving genes participating in neuronal RNA processing and alternative splicing. These are starting points, not treatments. Finding a genetic region associated with an illness does not mean scientists have discovered “the gene” responsible for it. Researchers still must determine what biological process is being affected, how that process influences illness and whether it can be safely altered.
The genetic analysis also pointed toward biological targets associated with a class of anti-migraine medications, consistent with previous genetic relationships between migraine and anorexia. But no one should read this study and conclude that migraine medication has been established as a treatment for anorexia. It has not.
The importance of such findings is that genetics may provide leads. Over time, those leads could help researchers identify biological pathways that might be targeted with new or repurposed medications. That possibility is especially important in anorexia, where medication development has lagged far behind many other serious psychiatric illnesses.
Could We Eventually Identify People at Risk Earlier?
Again possibly, but we are not there yet. Researchers can combine thousands of genetic differences into what is called a polygenic risk score. The goal is to estimate inherited susceptibility by looking at many small genetic signals rather than searching for a single causative gene.
The scores used in this study could statistically distinguish some people at greater risk, but their predictive power remains far too weak for routine clinical use. There is currently no useful genetic screening test for anorexia or binge eating. Parents should not buy commercial genetic tests expecting them to determine whether their child will develop an eating disorder, and clinicians should not make treatment decisions based on such testing.
The future could be different. Risk assessment might someday combine family history, genetic susceptibility, early eating behaviors, anxiety or compulsivity, metabolic characteristics, changes in growth, activity patterns, developmental factors and other warning signs. No single factor would have to predict an eating disorder. Together, they might eventually help clinicians identify some vulnerable people earlier, before the illness becomes deeply entrenched.
The Research is Important for Families
One consequence of biological research has nothing to do with technology. It concerns blame. Families were once directly blamed for causing eating disorders, particularly anorexia. Modern treatment has moved a considerable distance from those theories, but parents and loved ones can still carry enormous guilt.
Genetic research provides another reason to reject the idea that a parent simply caused a child’s anorexia by saying the wrong thing, serving the wrong food or creating the wrong family environment. At the same time, genetics should not create a different form of fatalism. A genetic vulnerability does not mean recovery is impossible.
Genes influence risk. They do not write an unavoidable script. A family history of heart disease does not guarantee that someone will die of heart disease, and genetic vulnerability to depression does not make depression untreatable. The same principle applies here. Biology can create vulnerability while nutrition, medical care, psychotherapy, family support, relationships and environment can still profoundly affect what happens. For families, perhaps the most important message is to take the illness seriously without treating it as the patient’s fault or anyone else’s.
What This Study Does Not Prove
The excitement surrounding this research requires boundaries. The study does not prove that eating disorders are “genetic diseases” in the simple sense that Huntington’s disease or cystic fibrosis are genetic diseases. Eating disorders involve many genes interacting with development and environment. Nor does it establish that sociocultural influences are irrelevant. A genetically vulnerable brain still develops in a social environment, and dieting, trauma, bullying, weight stigma, cultural expectations and other experiences can matter.
The study also does not establish that current treatments are misguided. Nutritional rehabilitation, psychotherapy, medical monitoring and family support remain central components of care. Nor does the research provide a clinically useful genetic test or a newly proven medication.
There are important limitations in who was studied. The principal genetic analyses involved people of European ancestry because sufficiently large datasets from other populations were not available, which limits how confidently the findings can be generalized to other populations. The samples were also overwhelmingly female, with approximately 96 percent of the binge-eating group and 94 percent of the anorexia group being female. The researchers found some indication that genetic risk identified in females also applies to males, but there were too few male participants to determine whether male eating disorders have distinctive genetic characteristics. That limitation matters in a field that has historically underdiagnosed and understudied boys and men.
Finally, the binge-eating analysis was broader than a study of formally diagnosed binge-eating disorder. Participants were identified through several datasets and definitions of binge-eating behavior. The findings therefore should not be described as though they apply exclusively to people with a formal diagnosis of binge-eating disorder.
What Changes Today?
Clinicians should not change treatment protocols because of this study alone. There is no new genetic test to order, no newly established medication to prescribe and no genetic result that tells a family which therapy will work. What should continue to change is our understanding of the illness.
The scientific evidence increasingly argues against viewing an eating disorder primarily as a problem of food, weight, body image or social influence with biology added as an afterthought. The emerging model is considerably broader. Eating disorders appear to involve complicated interactions among inherited vulnerability, brain function, psychiatric characteristics, metabolism, development and environment.
The visible behaviors remain critically important. Restriction must be interrupted. Malnutrition must be treated. Purging must be addressed. Binge eating must be understood and treated. Medical instability requires medical care. But those behaviors may be surface manifestations of biological systems that differ substantially among patients.
That possibility should change the questions we ask. Instead of only asking why a patient will not eat, we may eventually understand which biological systems make eating so threatening or difficult. Instead of asking why someone cannot stop binge eating, we may better understand how appetite, reward and behavioral control systems are interacting. Instead of wondering why treatment worked for one person and failed another, we may discover that the two patients never had biologically identical illnesses in the first place.
The old argument over whether eating disorders are psychological or biological may itself become increasingly obsolete. The brain is biological. Psychological experience can change biology. Genes influence behavior. Environment influences biology. Starvation changes the brain. Metabolism communicates with the brain. These systems interact.
A Different Way of Seeing the Patient
Perhaps the most immediate value of this research is not a medication, a genetic test or a new diagnostic category. It is perspective. Eating disorders can produce behaviors that bewilder experienced clinicians and terrify families. A patient may resist the very nutrition needed to keep them alive. Someone may repeatedly engage in behavior they desperately want to stop. Intelligence, love, logic and fear of consequences can all seem insufficient to overcome the illness.
Genetics does not explain every one of those experiences, but the expanding evidence increasingly tells us that we are looking at disease, not simply difficult behavior. The Nature Mental Health study is not the final answer to what causes anorexia or binge eating. It may eventually be remembered as one step in a much larger scientific transition, away from defining eating disorders almost entirely by what patients do and toward understanding the biological processes that make those behaviors so difficult to change.
That transition may ultimately matter far more than discovering any single gene. The real promise of genetics is not that medicine will someday tell a patient, “We found your eating disorder gene.” The more meaningful possibility is that medicine may someday be able to say something far more useful: we understand much more about the particular form of this illness affecting you, and because we understand it better, we know much more precisely how to treat it.














